- Email: [email protected]
Abrasive wear in ceramic laminated composites
Wear 260 (2006) 1104–1111
Abrasive wear in ceramic laminated composites G. de Portu a,d , L. Micele a,d,∗ , D. Prandstraller b , G. Palombarini b , G. Pezzotti c,d b
a Research Institute of Science and Technology for Ceramics, ISTEC-CNR, Via Granarolo 64, 48018 Faenza, Italy Institute of Metallurgy, Faculty of Industrial Chemistry, University of Bologna, Viale Risorgimento 4, 40136 Bologna, Italy c Ceramic Physics Laboratory, Kyoto Institute of Technology, KIT, Sakyo-ku, Matsugasaki, 606-8585 Kyoto, Japan d Research Institute for Nano-Science, RIN, Kyoto Institute of Technology, Kyoto, Japan
Received 26 January 2005; received in revised form 14 July 2005; accepted 28 July 2005 Available online 29 August 2005
Abstract Laminated ceramic structures in the system Al2 O3 /Al2 O3 + 3Y-TZP (A/AZ) were prepared using a tape casting technique in order to obtain ceramic layers with different compositions and thicknesses. Piezo-spectroscopy was used to evaluate the residual stresses arisen from a calibrated mismatch in thermal expansion coefficients of the layers during the sintering process of the composite. The dependence of the residual stresses in the A and AZ layers on their thickness ratio was established. A microscale ball cratering method was used to investigate the influence that the surface compressive stress can play on the abrasive wear resistance of the composite structures. The results were compared with those obtained with an unstressed reference material prepared either by lamination of pure alumina green-sheets or by cold isostatic pressing of alumina powder. The experimental results have shown that the abrasive wear resistance is higher for samples with compressive residual stresses within the surface regions. © 2005 Elsevier B.V. All rights reserved. Keywords: Laminated composites; Al2 O3 ; Abrasive wear; Residual stresses
1. Introduction Due to high hardness, high modulus and chemical inertness, ceramic materials are good candidates for tribological applications. It has been shown that the toughness also plays an important role in improving the wear resistance [1]. Consequently, an increase in toughness can lead to an improvement in the tribological behaviour of ceramics. Phase transformations induced at the surface by thermochemical treatments have been explored with the aim to increase the surface toughness of 3Y-TZP [2,3]. In this way, high fracture toughness at the surface was obtained but a worse tribological behaviour was observed [4]. On the other hand, it has been shown that laminated hybrid structures constituted by alternate layers of different materi∗
Corresponding author. Fax: +39 054646381. E-mail addresses: [email protected] (G.d. Portu), [email protected] (L. Micele), [email protected] (D. Prandstraller), [email protected] (G. Pezzotti). 0043-1648/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.wear.2005.07.009
als can be properly designed in order to induce a surface compressive stress leading to an improved surface toughness [5–8] and wear resistance [9]. Residual stresses arise from a mismatch between the coefficients of thermal expansion (CTE), sintering rates and elastic constants of the constituent phases and neighbouring layers [10,11]. Compressive residual stresses are induced in layers with lower CTE, while tensile stresses arise in those with higher CTE. However, it is worth noting that the residual stress field also depends on the geometry of the layered structure and in particular on the thickness ratio among layers [12–16]. The effectiveness of laminated hybrid structures in improving the sliding wear resistance of alumina has been already reported by Toschi et al. [9]. The aim of this work is to verify the influence that a surface compressive stress and then an increase in toughness can play on the resistance to abrasive wear of tape cast Al2 O3 in laminated structures, by means of a comparison with the resistance of unstressed alumina. For this purpose, three laminated composites were produced with ceramic layers differing in the thickness ratio, together with
G. de Portu et al. / Wear 260 (2006) 1104–1111
layered structures made of pure alumina and monolithic samples produced by powder sintering. The residual stresses were measured using a piezo-spectroscopic technique applied to the chromophoric fluorescence of Al2 O3 [15,16], while the abrasive wear resistance was determined using a microscale ball cratering method [17,18].
2. Experimental procedures 2.1. Specimen preparation Thin ceramic sheets were prepared by tape casting. Raw ceramic powders were Al2 O3 (Alcoa A16, Alcoa Aluminum Co., New York, USA) and ZrO2 (TZ3Y-S, Tosoh Corp., Japan), both with a mean particle size of 0.3 m. Sheets of pure alumina (hereinafter designated as A) as well as of the alumina–zirconia composite in the volume ratio 60/40 (hereinafter AZ) were produced. The CTE values are 9.0 × 10−6 K−1 for alumina and 10.0 × 10−6 K−1 for the alumina–zirconia composite [9]; this limited mismatch allows residual stresses to arise in the hybrid laminates in absence of microcracking or delamination effects [9]. The ceramic powders were ball mixed with solvents (ethanol and methyl-ethylketone), binders (polyvinylbutyral and dibuthylphtalate) and a surfactant (triolein), to prepare slurries suitable for tape casting. The green tapes were dried and punched into 34 mm × 50 mm components of rectangular shape. The stacking sequence of the components in the composite was designed in order to obtain three kinds of laminates with different AZ/A thickness ratio. The green-sheets were warm pressed at 80 ◦ C and at a pressure of 30 MPa for 30 min. Before sintering, an extremely careful burn out up to 600 ◦ C was carried out at low heating rate. Finally, the laminates were sintered at 1550 ◦ C for 1 h with heating and cooling rates of 30 ◦ C/h. In this way, hybrid laminates were produced with AZ/A thickness ratio equal to 1.12, 1.36 and 2.82 (S2, S3 and S4 samples, respectively). The composite S4 was prepared by stacking two AZ layers for each A layer. To obtain a symmetrical structure, two A layers were applied on each external side of the laminated structure. This also allowed one layer to be removed by grinding from each side of the sintered composite for a proper machining and surface finishing. As unstressed reference materials both a laminated structure (S1) and monolithic alumina (S0) were used. S1 contained 11 layers of pure alumina and were prepared with the same procedure and sintering conditions used for the hybrid composites, while S0 was prepared by cold isostatic pressing at 300 MPa using the same alumina powder used in tape casting. Sintering was carried out at the same temperature used for laminated composites. The density of the sintered samples was measured by the Archimede’s method. Both surfaces and cross-sections of both laminates and reference samples were ground and polished up to 1 m diamond paste in order to obtain optically
1105
flat surfaces, suitable for both abrasive wear tests and spectroscopic measurements. 2.2. Surface stress measurements The stress distribution along cross-sections of the multilayered samples was determined by a piezo-spectroscopic technique related to the characteristic R1, R2 doublet produced by the chromophoric fluorescence of Cr3+ impurities in Al2 O3 . The method of relating a line shift in a fluorescence spectrum to the stress state has been described by Grabner [19]. In polycrystalline, fine grained and untextured Al2 O3 samples subjected to a normal stress σ, a luminescence line undergoes a change in frequency ν given by the tensorial relationship: ν =
1 Πii σjj 3
(1)
where Π ii (the trace of piezo-spectroscopic matrix) is the piezo-spectroscopic coefficient relating frequency to stress. The spectrometric apparatus (T 64000 Horiba/JovinYvon) used had a 400 mW argon-ion laser beam operating at a wavelength of 488 nm as the excitation source. An optical microscope lens was used both to focus the laser beam on the sample and to collect the scattered signal, and a micron-scale magnification was obtained. The scattered frequencies were analysed with a triple monochromator equipped with a charge coupled device camera. When focussed by means of the optical microscope using a 20× optical lens, the laser spot on the sample was 5 m in size. Thermal and instrumental fluctuations were compensated by monitoring the spectrum from a Hg/Ne discharge lamp. The recorded spectra were analysed with a commercial software (LabSpec 4.02, Horiba/JobinIvon). The frequency shifts were obtained by subtracting the centre frequency of the peak, measured for the reference material in the unstressed state, from the centre frequency of the peak recorded under stress. A standard value of frequency for zero external stress was obtained acquiring an array of 100 spectra measured on the surface of sample S1 (produced by stacking only A layers) and averaging all the values of the peak centre. An important characteristic of the piezo-spectroscopic technique is that the average uniaxial piezo-spectroscopic coefficient Π uni , which characterises the linear dependence between peak shift and stress, strongly depends on several parameters which are specific of the material, and particularly on processing derived parameters such as grain size, presence of other phases, porosity, etc. Hence, a preliminar calibration procedure is required to determine the Π uni value pertinent to each material. For this purpose, bending bars were obtained from laminated structures prepared with layers of the same composition (A and AZ, respectively), mounted on a 4-points bending jig under the laser beam focussed using the optical microscope and loaded below the fracture stress. After loading, the jig was moved under the microscope and spectra were recorded every 40 m on going from the side in compression
1106
G. de Portu et al. / Wear 260 (2006) 1104–1111
towards the side in tension of the specimen. The load value was converted into a stress value, σ, using the standard 4point-bending elastic equation, and then the peak shift, ν, was plotted as a function of the applied stress. The average Π uni value was obtained from the slope of the σ versus ν plot. The stress distributions in the laminated structures were measured at a microscopic level by determining stress line profiles by means of automatic 10 m-spaced measurements carried out on the specimen cross-sections with the aid of a computerised X–Y table moved with a lateral resolution of 0.1 m along both axes.
The worn surfaces were examined by optical microscopy and scanning electron microscopy. The wear coefficient of each material was determined as the average of at least two series of tests by varying the sliding distance from 9 to 45 m. The depth of the wear craters was lower than 30 m. The wear volume was calculated measuring the diameter of the spherical crater with an optical microscope and using the following formulae:
2.3. Wear tests Tests on the resistance to abrasive wear were carried out following the method described elsewhere [17,18], using a commercial ball cratering apparatus illustrated schematically in Fig. 1 (Plint & Partners Ltd., UK, model TE66) which allows the normal load to be controlled with an accuracy of ±0.01 N. The ball is driven by a shaft, while the sample is vertically mounted on a pivoted L-shaped arm and is loaded against the ball by a dead weight hanging from the horizontal lever. A slurry containing SiC particles (7–8 m mean size, 70 g per 100 ml of distilled water) was used as the abrasive medium. The slurry was continuously stirred during the test to prevent sedimentation of SiC particles. The abrasive slurry was driven against the sample by a rotating steel sphere (25.4 mm diameter, and 990 ± 40 HV hardness). The ball rotation speed was 37.6 rpm in all tests, while the normal load ranged from 0.2 to 0.5 N. Before the use, the ball surface was appropriately treated to produce a fine pitting on the surface, a modification necessary to favour and make uniform the contact between abrasive particles and sample and to obtain consistent results.
Fig. 1. Schematic of microscale abrasion test set-up.
where D is the diameter of the ball (25.4 mm), d is the measured crater diameter and h is the calculated depth of the spherical crater produced by abrasion.
3. Results and discussion 3.1. Microstructure The density value almost reached the theoretical level for the bulk material S0 (ρrel = 99.1), while it was 98.3% for AA laminates S1 and 97.4 ± 0.2% for all the A/AZ hybrid laminates S2, S3, S4. As expected [20,21], a small amount of porosity was observed in sintered components due to the high content of organic substances used in the tape casting process. This porosity was almost equally distributed throughout the layers. Macroscopic and microscopic scale analysis carried out on the surface and cross section of the specimens evidenced the absence of both tunneling and delamination cracks. Fig. 2 shows a SEM micrograph of the interface between A and AZ layers. Alumina grains are the darker ones. A very sharp interface can be observed and along with a tendency of zirconia particles to segregate at the border with the alumina layer. The
Fig. 2. SEM micrograph of the A/AZ interface. On the right, the A layer (darker grains) and on the left the composite AZ.
G. de Portu et al. / Wear 260 (2006) 1104–1111
1107
Table 1 Salient piezo-spectroscopic characteristics of the investigated materials Material
A AZ
Π hydro (cm−1 /GPa)
R2
−6.55 ± 0.09 −9.0 ± 0.1
0.994 0.995
R1 position for unstressed Al2 O3 (cm−1 )
(nm)
14407.64 ± 0.02 14405.31 ± 0.02
694.076 694.188
was calculated as follows: σ=
Fig. 3. Dependence of the induced stress on the R1 band position for A and AZ multilayered composites. Confidence bands are shown as dashed lines.
grains of Al2 O3 in the composite are smaller than in the A layer as a consequence of grain growth hindering exerted by ZrO2 . However, all the abrasive tests were performed on pure alumina with comparable grain size in all samples, namely 3.1 ± 0.3, 2.9 ± 0.3 and 2.6 ± 0.2 m for S0, S1 reference materials and S2–S4 hybrid laminates, respectively. These differences are too small to affect to a significant extent the wear behaviour of the materials under examination.
3.2. Surface residual stresses Fig. 3 shows the dependence of the R1 band shift on the induced stress in the range from about −250 to +250 MPa. The slope of the least square linear regression was assumed as the uniaxial piezo-spectroscopic coefficient. Contrary to sapphire, which has a different value of the coefficient Π ii for each crystallographic axis [22], the Al2 O3 prepared according to the previously described process [8] is expected to be polycrystalline and untextured, so that the uniaxial coefficient is one of the three identical elements of the diagonal of the piezo-spectroscopic tensor Π ii . The hydrostatic stress
ν ν = Πhydro 3Πuni
(4)
where ν is the R1 peak shift. Using R1 band, Eq. (4) gives only the mean normal stress and no deviatoric components are present. In laminated structures, the cross-section is subject to a relaxation of hydrostatic stress known as “edge effect” [15], unless the measurements are carried out at a depth greater than the layer thickness. In our case, in the Al2 O3 layer, the depth probed by the laser was about 120–140 m [23]. For this reason, the edge effect could no be avoided and the residual stress measured with piezo-spectroscopy technique was lower than the actual bulk stress. No alternative methods were used to asses residual stresses; however, a comparison with previous measurements carried out using indentation technique [9] evidenced some differences. Nevertheless, because this systematic errors is similar for all the laminated samples, the effect of thickness ratio among the layers (i.e. of the amount of residual stress) on abrasive wear can be determined anyway. Table 1 summarises the salient piezo-spectroscopic characteristics of the materials under investigation. Eq. (4) allows the stress distribution through the laminates to be determined. Fig. 4 shows cross-sections of the S2, S3 and S4 A/AZ laminated structures and the residual stress profiles measured by fluorescence piezo-spectroscopy. In order to simplify the presentation, only half of the symmetrical stress profiles were plotted. The comparison among the profiles determined for the composites allows to point out that: (i) the average compressive stress in A layers increases with the increase in the AZ/A thickness ratio, and (ii) the residual stress within the surface regions decreases following a linear profile (Fig. 5). Both results are in agreement with previous
Fig. 4. Stress profiles determined for the multilayered composites S2, S3 and S4, each one referring to the adjacent cross-section. For simplicity, only one-half of the symmetric stress profile is plotted. For clarity, different scales were used for graphs and micrographs.
1108
G. de Portu et al. / Wear 260 (2006) 1104–1111 Table 2 Average values of surface stress in laminated composites calculated over the whole profile (centre column) and over the outermost 30 m thick region, interested by wear craters (right column) Composite
−35 ± 5 −52 ± 5 −185 ± 8
S2 S3 S4 a b
Whole profilea (MPa)
Region interested by wearb (MPa) −27 ± 4 −45 ± 4 −166 ± 7
This average refers to profiles shown in Fig. 5. This average refers to the first 30 m of profiles of Fig. 5.
3.3. Wear Fig. 5. Residual stress profiles within the outer layer of alumina on composites S2, S3, S4, as a function of the depth from the surface.
works on multilayered materials: in effect, the increase in the overall compressive stress in the layer with a lower CTE (A in the present case) as the AZ/A thickness ratio is increased has been already predicted by Lange et al. [12,13]. Others authors [15,16], in turn, have shown that a significant stress relaxation occurs on going towards the specimen surfaces free of constraint on one side. In Table 2 are shown the stress values obtained by averaging both the whole profile of the outer A layer and the stress values in the less than 30 m deep region interested by wear craters. The errors on the stress determination were calculated using both the average error on the peak centring and the error on the slope of calibration lines, estimated with Microcalc Origin 6.0.
A typical example of wear crater produced after the microscale abrasion tests is shown in the low magnification image (2.5×) reported in Fig. 6. In Fig. 6a is shown a sequence of spherical craters produced on S2 specimen by tests carried out with 0.2 N normal load for different sliding distances, while in Fig. 6b are shown the differences in size among craters produced on different materials using the same testing conditions (0.2 N normal load, 27 m sliding distance). No grooves were observed in the surface morphology of the craters produced on all samples which were analysed by scanning electron microscopy (Fig. 7a and b). This result indicates that the damage mechanism was typical of a three body abrasive wear. At higher magnification (Fig. 7c and d), weak plastic deformation and some very fine debris were observed in narrow areas of all samples. As shown by EDAX analyses, (i) no differences in composition exist between debris and surrounding base material, and (ii) the presence of embed-
Fig. 6. Optical micrograph of wear craters produced with 0.2 N applied load on: (a) S2 material for different sliding distances; (b) different materials for the same sliding distance (27 m).
G. de Portu et al. / Wear 260 (2006) 1104–1111
1109
Fig. 7. (a) SEM micrograph of a crater produced on the S2 material. (b–d) Morphological details of the scar at different magnifications.
ded SiC particles has to be excluded. Therefore, debris may form by abrasion of Al2 O3 , that is rubbed away by SiC particles but not completely removed by the slurry flux. The same behaviour has been shown by the different samples for all the experimental condition used.
Fig. 8. Wear volume determined as a function of sliding distance for bulk alumina (material S0, dashed line) and laminated structures (materials from S1 to S4, solid lines). The specific wear rate κ is given for all the materials by the slope of the corresponding straight line.
For abrasive wear of bulk materials, a simple model (equivalent to the Archard equation for sliding wear) can be used and leads to the following equation for the wear volume V: V =k×s×N
(5)
where s is the total sliding distance of the sphere relative to the specimen surface, N the normal load and κ the wear coefficient or specific wear rate. Fig. 8 shows the volume wear of each material tested under various experimental conditions (loads from 0.2 to 0.5 N, sliding distance from 21 to 45 m), plotted as a function of the product between sliding distance and normal load. The line corresponding to sample S0 (cold isostatically pressed and sintered Al2 O3 ) lies beneath the line referring to sample S1 (tape cast Al2 O3 ). The significantly lower resistance of alumina laminates is probably due to differences in the properties of materials prepared using different processes. In particular, porosity in tape cast materials from S1 to S4 is ∼2% higher than that measured for S0. However, the wear performance of tape cast materials can be recovered and further improved by the stimulation of suitable residual stresses. Lines referring to A/AZ lami-
1110
G. de Portu et al. / Wear 260 (2006) 1104–1111
Fig. 9. Wear coefficients κ for the same materials as in Fig. 8 determined as function of the residual stresses induced in the outermost 30 m regions (interested by wear craters), compared with the value of bulk alumina (material S0, dashed line). The error bars on κ were the confidence interval related to the slope of lines of Fig. 8 calculated with Microcalc Origin 6.0, whereas the bars related to the residual stress values were those of Table 2.
nated samples tested in abrasive wear regime under the same experimental condition clearly show how an increase in the surface compressive stresses (due to an increase in the AZ/A thickness ratio) leads to a lower wear volume. The presence of residual stresses of increasing intensity in the hybrid laminates S2, S3 and S4 allows to recover the performance deficit that, in the case of the laminated material S1, can be due to a higher porosity. The experimental data also show that for a laminated Al2 O3 /Al2 O3 –ZrO2 composite where the CTE mismatch between the layers is about 1.0 × 10−6 K−1 , the minimum value of the AZ/A thickness ratio, necessary to achieve a wear resistance comparable to that characterising the bulk material, is approximately 1. In Fig. 9, the specific wear coefficient is plotted versus the induced residual stress, in order to make the results independent of the composite geometry. A residual stress of about −35 MPa allows the hybrid laminates (with ∼2% higher porosity) to achieve the same abrasive wear resistance of bulk Al2 O3 . Due to the edge effect, afore mentioned, this threshold could be slightly higher; however, it appears clear that a stress of few tens of MPa on the surface is sufficient to improve the abrasive wear resistance of alumina. As known, the presence of compressive residual stresses increases the toughness of the surface layer [9]. Consequently, the propagation of a crack generated during the wear process can be considerably hindered to an extent increasing with the stress intensity. This phenomenon reduces the number of cracks which, reaching a critical size, become responsible for the material removal in form of wear debris.
4. Conclusions Laminated composites were produced by alternating ceramic layers of different composition and thickness allowing surface compressive stress states to establish, with the aim to obtain materials with improved toughness and wear
resistance. For this purpose, structures were prepared where layers of Al2 O3 and Al2 O3 + 3Y-TZP (A and AZ), selected on the basis of a calibrated difference between their coefficients of thermal expansion, and produced by tape casting, were alternated with different values of AZ/A thickness ratio. Compressive states of residual stresses were obtained along the cross-section of the external layers as the result of mismatch in CTE and differential shrinkage during the sintering process. Their intensity, evaluated by piezo-spectroscopic techniques, increased with the AZ/A ratio with a minimum at the outer surface and a profile linearly increasing on going inside within the layer. Correspondingly, the resistance to abrasive wear, evaluated by a ball cratering method, also increased with the AZ/A ratio becoming significantly higher than that displayed by samples of bulk Al2 O3 produced by cold isostatic pressing. The production of laminated structures with compressive residual stresses within the surface regions has been proved to be a suitable way to obtain composite materials with superior tribological properties.
Acknowledgements The authors gratefully acknowledge the Japan Society for the Promotion of Science (JSPS) and Italian National Research Council (CNR) for the financial support allowing G. de Portu to work in Japan, and the Italian Ministry for Foreign Affairs (MAE) for the support to the creation of Research Institute of Nanoscience.
References [1] T.E. Fischer, M.P. Anderson, S. Jahanmir, Influence of fracture toughness on the wear resistance of yttria-doped zirconium oxide, J. Am. Ceram. Soc. 72 (2) (1989) 252–257. [2] D.J. Green, A technique for introducing surface compression into zirconia ceramics, J. Am. Ceram. Soc. 66 (10) (1983) C-178–C-179. [3] G. de Portu, S. Conoci, Simplified equation for evaluating the influence of surface residual stresses on the toughness of zirconia ceramics, J. Am. Ceram. Soc. 80 (12) (1997) 3242–3244. [4] S. Conoci, C. Melandri, G. de Portu, Wear of Y-TZP containing compressive residual stresses at the surface, J. Mat. Sci. 34 (1999) 1009–1015. [5] R. Lakshminarayanan, D.K. Shetty, R.A. Cutler, Toughening of layered ceramic composites with residual surface compression, J. Am. Ceram. Soc. 79 (1) (1996) 79–87. [6] Kuo Dong-Hau, W.M. Kriven, A strong and damage-tolerant oxide laminate, J. Am. Ceram. Soc. 80 (9) (1997) 2421–2424. [7] P.Z. Cai, D.J. Green, G.L. Messing, Mechanical characterization of Al2 O3 /ZrO2 hybrid laminates, J. Eur. Ceram. Soc. 5 (1998) 2025–2034. [8] A. Tarlazzi, E. Roncari, P. Pinasco, S. Guicciardi, C. Melandri, G. de Portu, Tribological behaviour of Al2 O3 /ZrO2 –ZrO2 laminated composites, Wear 244 (2000) 29–40. [9] F. Toschi, C. Melandri, P. Pinasco, E. Roncari, S. Guicciardi, G. De Portu, Influence of residual stress on the wear behaviour of alumina/alumina-zirconia laminated composites, J. Am. Ceram. Soc. 86 (9) (2003) 1547–1553.
G. de Portu et al. / Wear 260 (2006) 1104–1111 [10] P.Z. Cai, D.J. Green, G.L. Messing, Constrained densification of alumina/zirconia hybrid laminates. I. Experimental observations of processing defects, J. Am. Ceram. Soc. 80 (8) (1997) 1929–1939. [11] P.Z. Cai, D.J. Green, G.L. Messing, Constrained densification of alumina/zirconia hybrid laminates. II. Viscoelastic stress computation, J. Am. Ceram. Soc. 80 (8) (1997) 1940–1948. [12] C. Hillman, Z. Suo, F.F. Lange, Cracking of laminates subjected to biaxial tensile stress, J. Am. Ceram. Soc. 79 (1996) 2127. [13] D.B. Marshall, J.J. Ratto, F.F. Lange, Enhanced fracture toughness in layered microcomposites of Ce–ZrO2 and Al2 O3 , J. Am. Ceram. Soc. 74 (1991) 2979. [14] T. Chartier, D. Merle, J.L. Besson, Laminar ceramic composites, J. Eur. Ceram. Soc. 15 (1995) 101–107. [15] V. Sergo, D.M. Lipkin, G. De Portu, D.R. Clarke, Edge stresses in alumina/zirconia laminates, J. Am. Ceram. Soc. 80 (7) (1997) 1633–1638. [16] G. de Portu, L. Micele, Y. Sekiguchi, G. Pezzotti, Measurements of residual stress distributions in Al2 O3 /3Y-TZP multilayered composites by fluorescence and Raman microprobe piezo-spectroscopy, Acta Materialia 53 (2005) 1511–1520. [17] R.I. Trezona, D.N. Allsopp, I.M. Hutchings, Transitions between two-body and three-body abrasive wear: influence of test conditions in the microscale abrasive wear test, Wear 225–229 (3) (1999) 205–214.
1111
[18] V. Imbeni, C. Martini, D. Prandstraller, G. Poli, C. Trepanier, T.W. Duerig, Preliminary study of micro-scale abrasive wear of a NiTi shape memory alloy, Wear 254 (2003) 1299–1306. [19] L. Grabner, Spectroscopic technique for the measurement of residual stress in sintered Al2 O3 , J. Appl. Phys. 49 (1978) 580. [20] C. Fiori, G. de Portu, Tape casting: a technique for preparing and studying new materials, in: R.W. Davidge (Ed.), British Ceramic Proceedings no. 38, Novel Ceramic Fabrication Processes and Applications, Shelton, Stoke-on-Trent, UK, December 1986, pp. 213– 225. [21] G. de Portu, C. Fiori, O. Sbaizero, Fabrication, microstructure and properties of ZrO2 —toughened Al2 O3 substrates, in: S. Somia, N. Yamamoto, H. Yanagida (Eds.), Advances in Ceramics: Science and Technology of Zirconia III, 24, The Am. Ceram. Soc. Inc., Columbus, OH, USA, 1988, pp. 1063–1073. [22] J. He, D.R. Clarke, Determination of the piezospectroscopic coefficients for chromium-doped sapphire, J. Am. Ceram. Soc. 78 (1995) 1347. [23] W. Zhu, G. Pezzotti, Spatially resolved stress analysis in Al2 O3 /3YTZP multilayered composites using confocal fluorescence spectroscopy, in: Masaaki Naka, Toshimi Yamane (Eds.), Proceedings of PSEA ’04, 2004, pp. 397–403.
Abrasive wear in ceramic laminated composites G. de Portu a,d , L. Micele a,d,∗ , D. Prandstraller b , G. Palombarini b , G. Pezzotti c,d b
a Research Institute of Science and Technology for Ceramics, ISTEC-CNR, Via Granarolo 64, 48018 Faenza, Italy Institute of Metallurgy, Faculty of Industrial Chemistry, University of Bologna, Viale Risorgimento 4, 40136 Bologna, Italy c Ceramic Physics Laboratory, Kyoto Institute of Technology, KIT, Sakyo-ku, Matsugasaki, 606-8585 Kyoto, Japan d Research Institute for Nano-Science, RIN, Kyoto Institute of Technology, Kyoto, Japan
Received 26 January 2005; received in revised form 14 July 2005; accepted 28 July 2005 Available online 29 August 2005
Abstract Laminated ceramic structures in the system Al2 O3 /Al2 O3 + 3Y-TZP (A/AZ) were prepared using a tape casting technique in order to obtain ceramic layers with different compositions and thicknesses. Piezo-spectroscopy was used to evaluate the residual stresses arisen from a calibrated mismatch in thermal expansion coefficients of the layers during the sintering process of the composite. The dependence of the residual stresses in the A and AZ layers on their thickness ratio was established. A microscale ball cratering method was used to investigate the influence that the surface compressive stress can play on the abrasive wear resistance of the composite structures. The results were compared with those obtained with an unstressed reference material prepared either by lamination of pure alumina green-sheets or by cold isostatic pressing of alumina powder. The experimental results have shown that the abrasive wear resistance is higher for samples with compressive residual stresses within the surface regions. © 2005 Elsevier B.V. All rights reserved. Keywords: Laminated composites; Al2 O3 ; Abrasive wear; Residual stresses
1. Introduction Due to high hardness, high modulus and chemical inertness, ceramic materials are good candidates for tribological applications. It has been shown that the toughness also plays an important role in improving the wear resistance [1]. Consequently, an increase in toughness can lead to an improvement in the tribological behaviour of ceramics. Phase transformations induced at the surface by thermochemical treatments have been explored with the aim to increase the surface toughness of 3Y-TZP [2,3]. In this way, high fracture toughness at the surface was obtained but a worse tribological behaviour was observed [4]. On the other hand, it has been shown that laminated hybrid structures constituted by alternate layers of different materi∗
Corresponding author. Fax: +39 054646381. E-mail addresses: [email protected] (G.d. Portu), [email protected] (L. Micele), [email protected] (D. Prandstraller), [email protected] (G. Pezzotti). 0043-1648/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.wear.2005.07.009
als can be properly designed in order to induce a surface compressive stress leading to an improved surface toughness [5–8] and wear resistance [9]. Residual stresses arise from a mismatch between the coefficients of thermal expansion (CTE), sintering rates and elastic constants of the constituent phases and neighbouring layers [10,11]. Compressive residual stresses are induced in layers with lower CTE, while tensile stresses arise in those with higher CTE. However, it is worth noting that the residual stress field also depends on the geometry of the layered structure and in particular on the thickness ratio among layers [12–16]. The effectiveness of laminated hybrid structures in improving the sliding wear resistance of alumina has been already reported by Toschi et al. [9]. The aim of this work is to verify the influence that a surface compressive stress and then an increase in toughness can play on the resistance to abrasive wear of tape cast Al2 O3 in laminated structures, by means of a comparison with the resistance of unstressed alumina. For this purpose, three laminated composites were produced with ceramic layers differing in the thickness ratio, together with
G. de Portu et al. / Wear 260 (2006) 1104–1111
layered structures made of pure alumina and monolithic samples produced by powder sintering. The residual stresses were measured using a piezo-spectroscopic technique applied to the chromophoric fluorescence of Al2 O3 [15,16], while the abrasive wear resistance was determined using a microscale ball cratering method [17,18].
2. Experimental procedures 2.1. Specimen preparation Thin ceramic sheets were prepared by tape casting. Raw ceramic powders were Al2 O3 (Alcoa A16, Alcoa Aluminum Co., New York, USA) and ZrO2 (TZ3Y-S, Tosoh Corp., Japan), both with a mean particle size of 0.3 m. Sheets of pure alumina (hereinafter designated as A) as well as of the alumina–zirconia composite in the volume ratio 60/40 (hereinafter AZ) were produced. The CTE values are 9.0 × 10−6 K−1 for alumina and 10.0 × 10−6 K−1 for the alumina–zirconia composite [9]; this limited mismatch allows residual stresses to arise in the hybrid laminates in absence of microcracking or delamination effects [9]. The ceramic powders were ball mixed with solvents (ethanol and methyl-ethylketone), binders (polyvinylbutyral and dibuthylphtalate) and a surfactant (triolein), to prepare slurries suitable for tape casting. The green tapes were dried and punched into 34 mm × 50 mm components of rectangular shape. The stacking sequence of the components in the composite was designed in order to obtain three kinds of laminates with different AZ/A thickness ratio. The green-sheets were warm pressed at 80 ◦ C and at a pressure of 30 MPa for 30 min. Before sintering, an extremely careful burn out up to 600 ◦ C was carried out at low heating rate. Finally, the laminates were sintered at 1550 ◦ C for 1 h with heating and cooling rates of 30 ◦ C/h. In this way, hybrid laminates were produced with AZ/A thickness ratio equal to 1.12, 1.36 and 2.82 (S2, S3 and S4 samples, respectively). The composite S4 was prepared by stacking two AZ layers for each A layer. To obtain a symmetrical structure, two A layers were applied on each external side of the laminated structure. This also allowed one layer to be removed by grinding from each side of the sintered composite for a proper machining and surface finishing. As unstressed reference materials both a laminated structure (S1) and monolithic alumina (S0) were used. S1 contained 11 layers of pure alumina and were prepared with the same procedure and sintering conditions used for the hybrid composites, while S0 was prepared by cold isostatic pressing at 300 MPa using the same alumina powder used in tape casting. Sintering was carried out at the same temperature used for laminated composites. The density of the sintered samples was measured by the Archimede’s method. Both surfaces and cross-sections of both laminates and reference samples were ground and polished up to 1 m diamond paste in order to obtain optically
1105
flat surfaces, suitable for both abrasive wear tests and spectroscopic measurements. 2.2. Surface stress measurements The stress distribution along cross-sections of the multilayered samples was determined by a piezo-spectroscopic technique related to the characteristic R1, R2 doublet produced by the chromophoric fluorescence of Cr3+ impurities in Al2 O3 . The method of relating a line shift in a fluorescence spectrum to the stress state has been described by Grabner [19]. In polycrystalline, fine grained and untextured Al2 O3 samples subjected to a normal stress σ, a luminescence line undergoes a change in frequency ν given by the tensorial relationship: ν =
1 Πii σjj 3
(1)
where Π ii (the trace of piezo-spectroscopic matrix) is the piezo-spectroscopic coefficient relating frequency to stress. The spectrometric apparatus (T 64000 Horiba/JovinYvon) used had a 400 mW argon-ion laser beam operating at a wavelength of 488 nm as the excitation source. An optical microscope lens was used both to focus the laser beam on the sample and to collect the scattered signal, and a micron-scale magnification was obtained. The scattered frequencies were analysed with a triple monochromator equipped with a charge coupled device camera. When focussed by means of the optical microscope using a 20× optical lens, the laser spot on the sample was 5 m in size. Thermal and instrumental fluctuations were compensated by monitoring the spectrum from a Hg/Ne discharge lamp. The recorded spectra were analysed with a commercial software (LabSpec 4.02, Horiba/JobinIvon). The frequency shifts were obtained by subtracting the centre frequency of the peak, measured for the reference material in the unstressed state, from the centre frequency of the peak recorded under stress. A standard value of frequency for zero external stress was obtained acquiring an array of 100 spectra measured on the surface of sample S1 (produced by stacking only A layers) and averaging all the values of the peak centre. An important characteristic of the piezo-spectroscopic technique is that the average uniaxial piezo-spectroscopic coefficient Π uni , which characterises the linear dependence between peak shift and stress, strongly depends on several parameters which are specific of the material, and particularly on processing derived parameters such as grain size, presence of other phases, porosity, etc. Hence, a preliminar calibration procedure is required to determine the Π uni value pertinent to each material. For this purpose, bending bars were obtained from laminated structures prepared with layers of the same composition (A and AZ, respectively), mounted on a 4-points bending jig under the laser beam focussed using the optical microscope and loaded below the fracture stress. After loading, the jig was moved under the microscope and spectra were recorded every 40 m on going from the side in compression
1106
G. de Portu et al. / Wear 260 (2006) 1104–1111
towards the side in tension of the specimen. The load value was converted into a stress value, σ, using the standard 4point-bending elastic equation, and then the peak shift, ν, was plotted as a function of the applied stress. The average Π uni value was obtained from the slope of the σ versus ν plot. The stress distributions in the laminated structures were measured at a microscopic level by determining stress line profiles by means of automatic 10 m-spaced measurements carried out on the specimen cross-sections with the aid of a computerised X–Y table moved with a lateral resolution of 0.1 m along both axes.
The worn surfaces were examined by optical microscopy and scanning electron microscopy. The wear coefficient of each material was determined as the average of at least two series of tests by varying the sliding distance from 9 to 45 m. The depth of the wear craters was lower than 30 m. The wear volume was calculated measuring the diameter of the spherical crater with an optical microscope and using the following formulae:
2.3. Wear tests Tests on the resistance to abrasive wear were carried out following the method described elsewhere [17,18], using a commercial ball cratering apparatus illustrated schematically in Fig. 1 (Plint & Partners Ltd., UK, model TE66) which allows the normal load to be controlled with an accuracy of ±0.01 N. The ball is driven by a shaft, while the sample is vertically mounted on a pivoted L-shaped arm and is loaded against the ball by a dead weight hanging from the horizontal lever. A slurry containing SiC particles (7–8 m mean size, 70 g per 100 ml of distilled water) was used as the abrasive medium. The slurry was continuously stirred during the test to prevent sedimentation of SiC particles. The abrasive slurry was driven against the sample by a rotating steel sphere (25.4 mm diameter, and 990 ± 40 HV hardness). The ball rotation speed was 37.6 rpm in all tests, while the normal load ranged from 0.2 to 0.5 N. Before the use, the ball surface was appropriately treated to produce a fine pitting on the surface, a modification necessary to favour and make uniform the contact between abrasive particles and sample and to obtain consistent results.
Fig. 1. Schematic of microscale abrasion test set-up.
where D is the diameter of the ball (25.4 mm), d is the measured crater diameter and h is the calculated depth of the spherical crater produced by abrasion.
3. Results and discussion 3.1. Microstructure The density value almost reached the theoretical level for the bulk material S0 (ρrel = 99.1), while it was 98.3% for AA laminates S1 and 97.4 ± 0.2% for all the A/AZ hybrid laminates S2, S3, S4. As expected [20,21], a small amount of porosity was observed in sintered components due to the high content of organic substances used in the tape casting process. This porosity was almost equally distributed throughout the layers. Macroscopic and microscopic scale analysis carried out on the surface and cross section of the specimens evidenced the absence of both tunneling and delamination cracks. Fig. 2 shows a SEM micrograph of the interface between A and AZ layers. Alumina grains are the darker ones. A very sharp interface can be observed and along with a tendency of zirconia particles to segregate at the border with the alumina layer. The
Fig. 2. SEM micrograph of the A/AZ interface. On the right, the A layer (darker grains) and on the left the composite AZ.
G. de Portu et al. / Wear 260 (2006) 1104–1111
1107
Table 1 Salient piezo-spectroscopic characteristics of the investigated materials Material
A AZ
Π hydro (cm−1 /GPa)
R2
−6.55 ± 0.09 −9.0 ± 0.1
0.994 0.995
R1 position for unstressed Al2 O3 (cm−1 )
(nm)
14407.64 ± 0.02 14405.31 ± 0.02
694.076 694.188
was calculated as follows: σ=
Fig. 3. Dependence of the induced stress on the R1 band position for A and AZ multilayered composites. Confidence bands are shown as dashed lines.
grains of Al2 O3 in the composite are smaller than in the A layer as a consequence of grain growth hindering exerted by ZrO2 . However, all the abrasive tests were performed on pure alumina with comparable grain size in all samples, namely 3.1 ± 0.3, 2.9 ± 0.3 and 2.6 ± 0.2 m for S0, S1 reference materials and S2–S4 hybrid laminates, respectively. These differences are too small to affect to a significant extent the wear behaviour of the materials under examination.
3.2. Surface residual stresses Fig. 3 shows the dependence of the R1 band shift on the induced stress in the range from about −250 to +250 MPa. The slope of the least square linear regression was assumed as the uniaxial piezo-spectroscopic coefficient. Contrary to sapphire, which has a different value of the coefficient Π ii for each crystallographic axis [22], the Al2 O3 prepared according to the previously described process [8] is expected to be polycrystalline and untextured, so that the uniaxial coefficient is one of the three identical elements of the diagonal of the piezo-spectroscopic tensor Π ii . The hydrostatic stress
ν ν = Πhydro 3Πuni
(4)
where ν is the R1 peak shift. Using R1 band, Eq. (4) gives only the mean normal stress and no deviatoric components are present. In laminated structures, the cross-section is subject to a relaxation of hydrostatic stress known as “edge effect” [15], unless the measurements are carried out at a depth greater than the layer thickness. In our case, in the Al2 O3 layer, the depth probed by the laser was about 120–140 m [23]. For this reason, the edge effect could no be avoided and the residual stress measured with piezo-spectroscopy technique was lower than the actual bulk stress. No alternative methods were used to asses residual stresses; however, a comparison with previous measurements carried out using indentation technique [9] evidenced some differences. Nevertheless, because this systematic errors is similar for all the laminated samples, the effect of thickness ratio among the layers (i.e. of the amount of residual stress) on abrasive wear can be determined anyway. Table 1 summarises the salient piezo-spectroscopic characteristics of the materials under investigation. Eq. (4) allows the stress distribution through the laminates to be determined. Fig. 4 shows cross-sections of the S2, S3 and S4 A/AZ laminated structures and the residual stress profiles measured by fluorescence piezo-spectroscopy. In order to simplify the presentation, only half of the symmetrical stress profiles were plotted. The comparison among the profiles determined for the composites allows to point out that: (i) the average compressive stress in A layers increases with the increase in the AZ/A thickness ratio, and (ii) the residual stress within the surface regions decreases following a linear profile (Fig. 5). Both results are in agreement with previous
Fig. 4. Stress profiles determined for the multilayered composites S2, S3 and S4, each one referring to the adjacent cross-section. For simplicity, only one-half of the symmetric stress profile is plotted. For clarity, different scales were used for graphs and micrographs.
1108
G. de Portu et al. / Wear 260 (2006) 1104–1111 Table 2 Average values of surface stress in laminated composites calculated over the whole profile (centre column) and over the outermost 30 m thick region, interested by wear craters (right column) Composite
−35 ± 5 −52 ± 5 −185 ± 8
S2 S3 S4 a b
Whole profilea (MPa)
Region interested by wearb (MPa) −27 ± 4 −45 ± 4 −166 ± 7
This average refers to profiles shown in Fig. 5. This average refers to the first 30 m of profiles of Fig. 5.
3.3. Wear Fig. 5. Residual stress profiles within the outer layer of alumina on composites S2, S3, S4, as a function of the depth from the surface.
works on multilayered materials: in effect, the increase in the overall compressive stress in the layer with a lower CTE (A in the present case) as the AZ/A thickness ratio is increased has been already predicted by Lange et al. [12,13]. Others authors [15,16], in turn, have shown that a significant stress relaxation occurs on going towards the specimen surfaces free of constraint on one side. In Table 2 are shown the stress values obtained by averaging both the whole profile of the outer A layer and the stress values in the less than 30 m deep region interested by wear craters. The errors on the stress determination were calculated using both the average error on the peak centring and the error on the slope of calibration lines, estimated with Microcalc Origin 6.0.
A typical example of wear crater produced after the microscale abrasion tests is shown in the low magnification image (2.5×) reported in Fig. 6. In Fig. 6a is shown a sequence of spherical craters produced on S2 specimen by tests carried out with 0.2 N normal load for different sliding distances, while in Fig. 6b are shown the differences in size among craters produced on different materials using the same testing conditions (0.2 N normal load, 27 m sliding distance). No grooves were observed in the surface morphology of the craters produced on all samples which were analysed by scanning electron microscopy (Fig. 7a and b). This result indicates that the damage mechanism was typical of a three body abrasive wear. At higher magnification (Fig. 7c and d), weak plastic deformation and some very fine debris were observed in narrow areas of all samples. As shown by EDAX analyses, (i) no differences in composition exist between debris and surrounding base material, and (ii) the presence of embed-
Fig. 6. Optical micrograph of wear craters produced with 0.2 N applied load on: (a) S2 material for different sliding distances; (b) different materials for the same sliding distance (27 m).
G. de Portu et al. / Wear 260 (2006) 1104–1111
1109
Fig. 7. (a) SEM micrograph of a crater produced on the S2 material. (b–d) Morphological details of the scar at different magnifications.
ded SiC particles has to be excluded. Therefore, debris may form by abrasion of Al2 O3 , that is rubbed away by SiC particles but not completely removed by the slurry flux. The same behaviour has been shown by the different samples for all the experimental condition used.
Fig. 8. Wear volume determined as a function of sliding distance for bulk alumina (material S0, dashed line) and laminated structures (materials from S1 to S4, solid lines). The specific wear rate κ is given for all the materials by the slope of the corresponding straight line.
For abrasive wear of bulk materials, a simple model (equivalent to the Archard equation for sliding wear) can be used and leads to the following equation for the wear volume V: V =k×s×N
(5)
where s is the total sliding distance of the sphere relative to the specimen surface, N the normal load and κ the wear coefficient or specific wear rate. Fig. 8 shows the volume wear of each material tested under various experimental conditions (loads from 0.2 to 0.5 N, sliding distance from 21 to 45 m), plotted as a function of the product between sliding distance and normal load. The line corresponding to sample S0 (cold isostatically pressed and sintered Al2 O3 ) lies beneath the line referring to sample S1 (tape cast Al2 O3 ). The significantly lower resistance of alumina laminates is probably due to differences in the properties of materials prepared using different processes. In particular, porosity in tape cast materials from S1 to S4 is ∼2% higher than that measured for S0. However, the wear performance of tape cast materials can be recovered and further improved by the stimulation of suitable residual stresses. Lines referring to A/AZ lami-
1110
G. de Portu et al. / Wear 260 (2006) 1104–1111
Fig. 9. Wear coefficients κ for the same materials as in Fig. 8 determined as function of the residual stresses induced in the outermost 30 m regions (interested by wear craters), compared with the value of bulk alumina (material S0, dashed line). The error bars on κ were the confidence interval related to the slope of lines of Fig. 8 calculated with Microcalc Origin 6.0, whereas the bars related to the residual stress values were those of Table 2.
nated samples tested in abrasive wear regime under the same experimental condition clearly show how an increase in the surface compressive stresses (due to an increase in the AZ/A thickness ratio) leads to a lower wear volume. The presence of residual stresses of increasing intensity in the hybrid laminates S2, S3 and S4 allows to recover the performance deficit that, in the case of the laminated material S1, can be due to a higher porosity. The experimental data also show that for a laminated Al2 O3 /Al2 O3 –ZrO2 composite where the CTE mismatch between the layers is about 1.0 × 10−6 K−1 , the minimum value of the AZ/A thickness ratio, necessary to achieve a wear resistance comparable to that characterising the bulk material, is approximately 1. In Fig. 9, the specific wear coefficient is plotted versus the induced residual stress, in order to make the results independent of the composite geometry. A residual stress of about −35 MPa allows the hybrid laminates (with ∼2% higher porosity) to achieve the same abrasive wear resistance of bulk Al2 O3 . Due to the edge effect, afore mentioned, this threshold could be slightly higher; however, it appears clear that a stress of few tens of MPa on the surface is sufficient to improve the abrasive wear resistance of alumina. As known, the presence of compressive residual stresses increases the toughness of the surface layer [9]. Consequently, the propagation of a crack generated during the wear process can be considerably hindered to an extent increasing with the stress intensity. This phenomenon reduces the number of cracks which, reaching a critical size, become responsible for the material removal in form of wear debris.
4. Conclusions Laminated composites were produced by alternating ceramic layers of different composition and thickness allowing surface compressive stress states to establish, with the aim to obtain materials with improved toughness and wear
resistance. For this purpose, structures were prepared where layers of Al2 O3 and Al2 O3 + 3Y-TZP (A and AZ), selected on the basis of a calibrated difference between their coefficients of thermal expansion, and produced by tape casting, were alternated with different values of AZ/A thickness ratio. Compressive states of residual stresses were obtained along the cross-section of the external layers as the result of mismatch in CTE and differential shrinkage during the sintering process. Their intensity, evaluated by piezo-spectroscopic techniques, increased with the AZ/A ratio with a minimum at the outer surface and a profile linearly increasing on going inside within the layer. Correspondingly, the resistance to abrasive wear, evaluated by a ball cratering method, also increased with the AZ/A ratio becoming significantly higher than that displayed by samples of bulk Al2 O3 produced by cold isostatic pressing. The production of laminated structures with compressive residual stresses within the surface regions has been proved to be a suitable way to obtain composite materials with superior tribological properties.
Acknowledgements The authors gratefully acknowledge the Japan Society for the Promotion of Science (JSPS) and Italian National Research Council (CNR) for the financial support allowing G. de Portu to work in Japan, and the Italian Ministry for Foreign Affairs (MAE) for the support to the creation of Research Institute of Nanoscience.
References [1] T.E. Fischer, M.P. Anderson, S. Jahanmir, Influence of fracture toughness on the wear resistance of yttria-doped zirconium oxide, J. Am. Ceram. Soc. 72 (2) (1989) 252–257. [2] D.J. Green, A technique for introducing surface compression into zirconia ceramics, J. Am. Ceram. Soc. 66 (10) (1983) C-178–C-179. [3] G. de Portu, S. Conoci, Simplified equation for evaluating the influence of surface residual stresses on the toughness of zirconia ceramics, J. Am. Ceram. Soc. 80 (12) (1997) 3242–3244. [4] S. Conoci, C. Melandri, G. de Portu, Wear of Y-TZP containing compressive residual stresses at the surface, J. Mat. Sci. 34 (1999) 1009–1015. [5] R. Lakshminarayanan, D.K. Shetty, R.A. Cutler, Toughening of layered ceramic composites with residual surface compression, J. Am. Ceram. Soc. 79 (1) (1996) 79–87. [6] Kuo Dong-Hau, W.M. Kriven, A strong and damage-tolerant oxide laminate, J. Am. Ceram. Soc. 80 (9) (1997) 2421–2424. [7] P.Z. Cai, D.J. Green, G.L. Messing, Mechanical characterization of Al2 O3 /ZrO2 hybrid laminates, J. Eur. Ceram. Soc. 5 (1998) 2025–2034. [8] A. Tarlazzi, E. Roncari, P. Pinasco, S. Guicciardi, C. Melandri, G. de Portu, Tribological behaviour of Al2 O3 /ZrO2 –ZrO2 laminated composites, Wear 244 (2000) 29–40. [9] F. Toschi, C. Melandri, P. Pinasco, E. Roncari, S. Guicciardi, G. De Portu, Influence of residual stress on the wear behaviour of alumina/alumina-zirconia laminated composites, J. Am. Ceram. Soc. 86 (9) (2003) 1547–1553.
G. de Portu et al. / Wear 260 (2006) 1104–1111 [10] P.Z. Cai, D.J. Green, G.L. Messing, Constrained densification of alumina/zirconia hybrid laminates. I. Experimental observations of processing defects, J. Am. Ceram. Soc. 80 (8) (1997) 1929–1939. [11] P.Z. Cai, D.J. Green, G.L. Messing, Constrained densification of alumina/zirconia hybrid laminates. II. Viscoelastic stress computation, J. Am. Ceram. Soc. 80 (8) (1997) 1940–1948. [12] C. Hillman, Z. Suo, F.F. Lange, Cracking of laminates subjected to biaxial tensile stress, J. Am. Ceram. Soc. 79 (1996) 2127. [13] D.B. Marshall, J.J. Ratto, F.F. Lange, Enhanced fracture toughness in layered microcomposites of Ce–ZrO2 and Al2 O3 , J. Am. Ceram. Soc. 74 (1991) 2979. [14] T. Chartier, D. Merle, J.L. Besson, Laminar ceramic composites, J. Eur. Ceram. Soc. 15 (1995) 101–107. [15] V. Sergo, D.M. Lipkin, G. De Portu, D.R. Clarke, Edge stresses in alumina/zirconia laminates, J. Am. Ceram. Soc. 80 (7) (1997) 1633–1638. [16] G. de Portu, L. Micele, Y. Sekiguchi, G. Pezzotti, Measurements of residual stress distributions in Al2 O3 /3Y-TZP multilayered composites by fluorescence and Raman microprobe piezo-spectroscopy, Acta Materialia 53 (2005) 1511–1520. [17] R.I. Trezona, D.N. Allsopp, I.M. Hutchings, Transitions between two-body and three-body abrasive wear: influence of test conditions in the microscale abrasive wear test, Wear 225–229 (3) (1999) 205–214.
1111
[18] V. Imbeni, C. Martini, D. Prandstraller, G. Poli, C. Trepanier, T.W. Duerig, Preliminary study of micro-scale abrasive wear of a NiTi shape memory alloy, Wear 254 (2003) 1299–1306. [19] L. Grabner, Spectroscopic technique for the measurement of residual stress in sintered Al2 O3 , J. Appl. Phys. 49 (1978) 580. [20] C. Fiori, G. de Portu, Tape casting: a technique for preparing and studying new materials, in: R.W. Davidge (Ed.), British Ceramic Proceedings no. 38, Novel Ceramic Fabrication Processes and Applications, Shelton, Stoke-on-Trent, UK, December 1986, pp. 213– 225. [21] G. de Portu, C. Fiori, O. Sbaizero, Fabrication, microstructure and properties of ZrO2 —toughened Al2 O3 substrates, in: S. Somia, N. Yamamoto, H. Yanagida (Eds.), Advances in Ceramics: Science and Technology of Zirconia III, 24, The Am. Ceram. Soc. Inc., Columbus, OH, USA, 1988, pp. 1063–1073. [22] J. He, D.R. Clarke, Determination of the piezospectroscopic coefficients for chromium-doped sapphire, J. Am. Ceram. Soc. 78 (1995) 1347. [23] W. Zhu, G. Pezzotti, Spatially resolved stress analysis in Al2 O3 /3YTZP multilayered composites using confocal fluorescence spectroscopy, in: Masaaki Naka, Toshimi Yamane (Eds.), Proceedings of PSEA ’04, 2004, pp. 397–403.